This Article Appeared in a Journal Published by Elsevier. the Attached

This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright Author's personal copy

Opinion

Urgent preservation of boreal carbon stocks and biodiversity

Corey J.A. Bradshaw1,2*, Ian G. Warkentin3* and Navjot S. Sodhi4,5*

1 The Environment Institute and School of Earth & Environmental Sciences, University of Adelaide, South Australia 5005, Australia 2 South Australian Research and Development Institute, P.O. Box 120, Henley Beach, South Australia 5022, Australia 3 Environmental Science – Biology, Memorial University of Newfoundland, Corner Brook, Newfoundland and Labrador A2H 6P9, Canada 4 Department of Biological Sciences, National University of Singapore, 14 Science Drive 4, Singapore 117543, Republic of Singapore 5 Department of Organismic and Evolutionary Biology, Harvard University, 26 Oxford Street, Cambridge, MA 02138, USA

Containing approximately one-third of all remaining ment and climate change [8]. Although less immediately global forests, the boreal ecosystem is a crucial store threatened by deforestation than the tropics, these remain- of carbon and a haven for diverse biological commu- ing havens of the boreal forest could quickly become as nities. Historically, fire and insects primarily drove the threatened as tropical systems [1] while releasing substan- natural dynamics of this biome. However, human- tial amounts of C into the atmosphere [9]. mediated disturbances have increased in these forests Based on our review of data on the changes occurring in during recent years, resulting in extensive forest loss for boreal forest cover, we propose here that immediate action some regions, whereas others face heavy forest frag- must be taken to preserve this vital world resource. mentation or threat of exploitation. Current management practices are not likely to maintain the attendant A rapidly changing forest boreal forest communities, nor are they adequate to We used the Boreal Forest Monitoring Project’s [10] deli- mitigate climate change effects. There is an urgent need neation of the region (2000 to 2005) to assess patterns of to preserve existing boreal forests and restore degraded change over time. They defined the boreal zone based on areas if we are to avoid losing this relatively intact the Terrestrial Ecoregions map of the World Wildlife Fund biodiversity haven and major global carbon sink. [11], with modifications to add ecoregions of temperate coniferous and mixed forests characterized by similar sea- Introduction sonality and presence of winter snow cover. Also included Much world attention has focused on the loss and degra- were forested areas of forest-steppe ecoregions within dation of tropical forests over the last three decades [1].An continental North America and Asia and those along for- expansive reservoir for global biodiversity, these forests est-tundra transitional ecoregions; excluded due to data also contain substantial stores of terrestrial carbon (C) and limitations were small portions of boreal forest in Iceland have an enormous influence on regional and global cli- and regions >708 N latitude in Siberia. mates through evaporative cooling processes and the The second largest biome in the world, the circumpolar sequestration of C linked to high primary productivity boreal forest represents 32–33% of all the Earth’s forests [2]. Although concern rightly persists over continued [12], of which 22% is found in Russia alone (78% of this is in exploitation of tropical forests [1], a more global perspect- Siberia and the remainder in European Russia) ive on forest loss is necessary so that growing threats to (Figure 1a). The other five countries housing the remaining other ecosystems are not ignored [3]. Constituting about majority of boreal forest are Canada, USA, Sweden, Fin- one-third of extant forests on Earth and home to nearly land and Norway, although there are some large areas of half of the remaining large tracts of intact forest, boreal boreal forest in northern Mongolia and north-eastern ecosystems support a diverse flora and fauna and likewise China (Figure 1a). In 2005, an estimated 31% of all remain- harbour a substantial portion of global C stocks [4]. ing primary forests (‘forests of native species, in which Human populations are typically sparse in boreal zones there are no clearly visible indications of human activity so there has been relatively limited resource exploitation in and ecological processes are not significantly disturbed’) these areas, and disturbance dynamics have been largely worldwide were found in Russia and Canada alone [13].An driven by natural processes such as fire [5]. Consequently, estimated 80% of Canada’s boreal forest is thought to be few regions of the boreal forest have been extensively unfragmented by human settlements and roads [7]. modified compared with their tropical counterparts [6]. Nonetheless, fragmentation in boreal forests is on the However, rising demand for resources (mineral, energy, rise. The World Intact Forest Landscapes assessment [14] timber) has increased the extent of perturbation [7], while paints a dismal picture of the intactness of this biome fire dynamics have been altered due to human encroach- (Figure 1b). Using the definition of ‘intact’ as ‘areas 2 Corresponding author: Warkentin, I.G. ([email protected]) 500 km , internally undivided by infrastructure (e.g. * All authors contributed equally to this manuscript.. roads) and with linear dimensions 10 km’ (see Online

Opinion Trends in Ecology and Evolution Vol.24 No.10

Figure 1. (a) Global extent and major classification (deciduous broadleaf, mixed, evergreen needle-leaf, shrubland, and disturbed, burnt, urban or cropland) of boreal forests according to the 2000 Global Landcover dataset (www-tem.jrc.it/glc2000/) [78,79]. The boreal zone limits are taken from the Boreal Forest Monitoring (2000–2005) project [10] (http://globalmonitoring.sdstate.edu/projects/gfm/). The classification ‘shrubland’ includes broadleaved evergreen, deciduous, needle-leaved and dwarf shrubland, and grassland with sparse tree or shrub layers [78,79]. (b) Area of ‘intact’ boreal forest as defined by the World Intact Forest Landscapes assessment [14]. ‘Intact’ forest patches are 500 km2 in area that are internally undivided by infrastructure (e.g. roads) and with linear dimensions 10 km. (c) Tree canopy density measured as % coverage per pixel (0–100% colour gradient; 500 m resolution) from the 2000 MODIS-based Vegetation Continuous Field layers [80]. This dataset was used to make the World Intact Forest Landscapes assessment [14] shown in (b). The map projection for all panels is North Pole Lambert Azimuthal Equal Area. supplementary material), the ecologically contiguous areas fragments [12] due to: (i) increasing threats from logging; of the boreal forest cover only 44% of the biome (ii) rapid urban development; (iii) deciduous regrowth; (iv) (Figure 1b, c). Although the deﬁnition of ‘intactness’ is dam construction; (v) peat and other mining; and (vi) arbitrary, the maintenance of large habitat areas is necess- increasing frequency of ﬁres. ary to ensure the persistence of most species in a landscape According to the United Nations Temperate and Boreal [15]. The large extent of boreal forests belies the increasing Forest Resources Assessment 2000 (TBFRA) [16] and the fragmentation occurring there. In our opinion, the decreas- Food and Agriculture Organization of the United Nations’ ing quality of boreal forests should be a cause of concern [1]. Global Forest Assessment 2005 (GFA) [13], measures of The world’s most expansive and once contiguous forest forest extent in Canada and Fennoscandia have changed in Russia is rapidly turning into a network of smaller little in recent years [12]. However, even though elsewhere

542 Author's personal copy

Opinion Trends in Ecology and Evolution Vol.24 No.10 there is a perception that boreal forests and temperate forests are increasing in area from past deforestation, the forests are decreasing in ‘quality’ (see definitions in Online supplementary material) [14,16]. Apart from Russia, the proportion of ‘undisturbed’ forest in each of the six major boreal countries is <30% (Figure 2), with most (>60%) of forests in the USA and Scandinavia considered ‘semi- natural’ (see definitions in Online supplementary material) [16]. Another concern is that the total area considered unavailable for wood supply (i.e. ‘areas protected in strict nature reserves, wilderness areas, national parks, national monuments, habitat and species management areas, protected landscapes, or managed resource protection areas’) [16] does not exceed 2.5 Â 107 ha in any country, representing an area <10% of the total forested land in all boreal countries except Sweden (the latter has 20% protected) (Figure 2). Data from the TBFRA show that Russia holds the record among the 55 temperate and boreal countries and boreal countries assessed for the greatest annual decline in forest area (1.1 M ha year–1 between 1988 and 1993). Despite the decline, Russia still contains nearly 9 Â 108 ha (9 Â 106 km2) of ‘forest’ (defined as having tree crown cover >10% and a minimum patch area >0.5 ha) and ‘other wooded land’ (OWL; defined as having tree crown cover 5–10%) (Figure 2, see Online supplementary material), of which >80% of forest and OWL combined is considered ‘undisturbed’ by human activities [16]. Canada has the next largest area of undisturbed forest, but considerably less than Russia (nearly 90% less) at slightly more than 1 Â 108 ha that has changed little in extent since 1990 [13]. The USA has <2 Â 107 ha of boreal forest (mainly in Alaska), and each of the Fennoscandian boreal countries has <4.5 Â 106 ha (Figure 2). Fire has been the major disturbance process operating in boreal forests since the last Ice Age, [17] mainly because human population density is relatively low in boreal areas compared with most of the other biomes in the world [4]. However, advancing timber harvest (logging) and other Figure 2. (a) Total forest cover in the six main boreal countries categorized as human encroachment has led to an increase in fire fre- ‘forest’ (tree crown cover with >10% and area >0.5 ha) and ‘other wooded land’ (OWL = tree crown cover 5–10%), and whether or not ‘undisturbed’ by human quency in recent years, particularly in Siberia [12]. For activities according to the United Nations Temperate and Boreal Forest Resources example, in Russia, an area of 7.5 M ha burnt in 2002 and Assessment 2000 (TBFRA) [16]. (b) Values expressed as percentages for each 14.5 M ha burnt in 2003 [18], of which most (87% between country [16]. See Online Supplementary Material for more detailed definitions of forest types. (c) Total area and proportion of total forest (tree crown cover with 2002 and 2005) was started by humans [8]. This is com- >10% and area >0.5 ha) not available for wood supply according to the TBFRA pared with a annual mean burning rate of <4.5 M ha since 2000 [16] for each of the six main boreal countries. The TBFRA definition of ‘unavailable for wood supply’ is taken from the IUCN as ‘an area of land and/or sea the 1950s, which is more an index of the long-term natural especially dedicated to the protection and maintenance of biological diversity; and burning rate [17,19] (Online supplementary material of natural and associated cultural resources, and managed through legal or other Figure S1). In Russia in particular, most fires occur near effective means’. roads and other transportation networks, indicating that humans have a constant multiplication effect on fire events (up to eight-times above background rates) [8,17]. Weather Pacific Rim) [13,16]. For birds and mammals in the boreal anomalies related to human-driven climate change appear forest zone, the lowest diversity at all taxonomic levels to have increased fire susceptibility in recent years [8], but occurs in Europe, and the highest in western North Amer- humans are directly responsible for most ignitions in non- ica and east Asia [20]. To examine the degree to which intact Russian forests, and from 72% to 78% of ignitions in boreal species are threatened, we searched the IUCN’s all boreal forest types combined [8,17]. 2008 Red List (www.iucnredlist.org) for species under the ‘boreal forest’ habitat heading, which listed 367 Biodiversity threats species. After removing some mistakenly classified non- There are currently about 20300 species found within the boreal species and verifying distributions, there were boreal forest zone [4], but tree diversity in this zone is 348 species in the Red List in this category. Of these, relatively low compared with other temperate forests (e.g. >94% were listed as Least Concern – the remainder were

543 Author's personal copy

Opinion Trends in Ecology and Evolution Vol.24 No.10

Table 1. Taxonomic breakdown (by percentage) of threatened Earth, with an estimated 550 Gt C in combined soil and a species above-ground pools [21]. Although the boreal forest has Taxon IUCN Red List 2008b TBFRA 2000c,d primarily been considered a long-term global C sink, recent Fungi and lichens 515.9studies suggest that the rate of uptake may not be as high Plants 5 – as once thought [22]. Various models additionally predict Ferns – 1.4 Mosses – 16.1 that the boreal biome is the region most likely to be altered Vascular plantse 23.0 by climate change over the next century, with warmer Trees 1.6 temperatures and longer growing seasons [23] shifting it Butterflies and moths – 20.1 from being a net C sink to a source [9]. This warming is also Birds 50 7.6 predicted to lead to boreal forest expansion northwards Mammals 35 5.2 and upwards in elevation, whereas southern regions may Other vertebratesf 5 9.1 shift to grassland or temperate zone forest types [24]. aSee also Online Supplementary Material text and Tables S3 and S5 for more detail. bIUCN 2008 Red List (www.iucnredlist.org). There were 348 entries designated as Changing tree lines would alter albedo because former ‘boreal forest’ species, of which 328 (94.3%) fall into the Least Concern category (not winter snow-covered areas would then have decreased threatened); percentages therefore summarize the remaining 20 threatened species reflectivity, but warmer temperatures would also decrease by taxon. cUnited Nations Temperate and Boreal Forest Resources Assessment 2000 (TBFRA) the extent of snowfall and potentially lessen this impact [16]. The percentages are derived from the among-country (Canada, USA, Norway, [2]. Chen et al. [25] found that warmer forests had greater Sweden, Finland and Russia) mean number of ‘endangered’ species (1927 species listed in total; average of 321 endangered species per country) per taxonomic C sequestration, but others suggest that longer growing category. seasons and warmer temperatures are more likely to lead dSum of among-country mean percentages does not equal 100%. to greater decomposition rates [26]. An increased concen- eExcluding trees. fReptiles and amphibians (IUCN); reptiles, amphibians and fish (TBFRA). tration of CO2 and resulting fertilization could enhance boreal forest productivity [27], but this may not compen- considered Threatened (taxonomic breakdown in Table 1, sate for other negative influences of climate change on C and IUCN category breakdown and listed threatened sequestration processes, nor would the extent be equal species in Online supplementary material Tables S3 and across the globe [28]. S4). Estimates of the spatial distribution of C stores within We also examined the TBFRA 2000 database [16] for regions are not well-developed and can be highly variable additional data on species endangerment patterns in the [29]. Likewise, improved technological capacity has six main boreal countries. The TBFRA compiled infor- resulted in lowered estimates of C stocks for boreal forests mation based on IUCN Red List data, national compendia in recent years [30]. Combined, this uncertainty has caused and taxonomically specific lists and, although incomplete, concern over the validity of some global C modelling pro- now out-of-date, and including some temperate (not jections [31]. Clearly, Russia contains the largest area of strictly ‘boreal’) species that may bias the number of boreal forest of any country (Figures 1 and 2) and, by threatened species upward, the dataset probably extrapolation, potentially the most extensive C stores, represents a more detailed assessment of biodiversity but some analyses suggest that the C stock is not pro- trends in this region compared to the 2008 Red List portional to forest area [29,30]. Nevertheless, the accuracy because the TBFRA includes expert opinion and national of these estimates is hindered by discrepancies not only in listings for many species not adequately assessed by general methods, but also because of uncertainty in basic the IUCN [16]. The mean proportions of ‘endangered’ forest estimates of forest cover and which parts of the C pool species (i.e. assessed generally as of conservation concern) should be included [31] (Online supplementary material within nine taxonomic groups are listed in Table 1 (a Table S6). summary by boreal country is presented in Online supplementary material Table S5). Across all taxa, Fennoscan- Role of fire dian countries have the highest proportional Primary among the drivers of boreal forest dynamics and endangerment, with Sweden exceeding all other countries the associated C flux is fire [32]; its role in successional for all taxa except trees and ‘other vertebrates’ (reptiles, processes shapes the local and regional age structure and amphibians and fish); these last two categories are highest tree species composition of stands, thus influencing C for Finland and Norway, respectively (Online supple- sequestration patterns. Although high-intensity, stand- mentary material Table S5). Even though categorization replacing fires are more frequent in the North American of ‘forest-occurring’ species (see Online supplementary boreal zone than in the Siberian boreal zone [33], changing material) was not always available, and assessments can climate and weather patterns over the past 50 years have be more conservative in some countries, the results gener- altered fire dynamics and the release rates of C throughout ally appear to reflect the distribution of highest human the circumpolar region [5]. More frequent fires have been population density and development relative to total forest associated with increasing frequency of temperature area. Nonetheless, the boreal forest is a critical habitat for anomalies and more human-ignited fires [8,17], and a its threatened species, whereas its migratory species also 74–118% increase in the area burnt annually across North need simultaneous preservation in tropical areas. America is predicted over the next 100 years [34].Asa result, increased annual rates of C release have been Changing patterns of carbon storage and flux predicted [35] and indeed, forest fires in the boreal zone Like its proportional forest coverage, the boreal ecosystem have released more C over time: fire emissions of total C in contains roughly 30% of the stored terrestrial C of the North America approximately doubled from around 30 Tg

544 Author's personal copy

Opinion Trends in Ecology and Evolution Vol.24 No.10

C year–1 in the 1960s to >60 Tg C yearÀ1 in the 1990s, and can increase the frequency, intensity [17] and type (e.g. in Eurasia from 100–200 Tg C year–1 in 1996–1997 to shifting dominance of surface fires to crownfires [48]) nearly 500 Tg C yearÀ1 in 2002 [36]. The impact that C of fires. Natural disturbance is the pervasive force released by fire has on overall greenhouse gas concen- shifting C storage patterns in the boreal forest [32],and trations is an important component when trying to although management strategies may need to be modified determine whether boreal forests are a net sink or net to accommodate climate change [2,24], it is plausible that source of C. shifts in harvest management itself will have a limited effect [37]. Role of insect outbreaks The interactive effect of increasing fire frequency and Insect infestations also appear to exert a cyclical (but insect outbreaks arising from warmer temperatures, and a strong) influence on boreal forest C dynamics. Canada’s changing structure and composition of forests resulting boreal zone has recently shifted from a C sink in the 1990s from broad-scale harvest management, appears poised to to a C source in 2001 as warmer temperatures reduced lower boreal C stores and increase C emissions in the over-winter mortality of tree-killing insects, resulting in an foreseeable future [35–37,49]. However, some models increased frequency and severity of outbreaks and sub- suggest that the higher albedo of deforested areas covered sequent mass tree mortality [37]. Although some suggest in snow [50] provides a cooling effect that more than offsets that these outbreak dynamics are currently within the the warming associated with the release of CO2 through normal range for forest insects [5], an increased frequency deforestation [51,52]. Indeed, the heat retained by intact and severity of insect disturbances is anticipated as war- boreal forests contributes more to increased mean annual mer and potentially drier weather enables range expan- global temperature than any other biome [53]. It is our sions and intensification of outbreaks [37–39]. Some opinion that avoidance of deforestation and the mainten- evidence indicates that a higher incidence and severity ance of the boreal zone as a net C sink will provide more of insect outbreaks is already taking place, [e.g. 40] with durable climate warming mitigation (and obviously better the potential for sequential outbreaks by species such as prospects for biodiversity maintenance) because the ampli- the mountain pine beetle (Dendroctonus ponderosae) and fication effect (i.e. increasing temperatures melt more spruce budworm (Choristoneura fumiferana) causing snow, decreasing surface albedo and raising temperatures Canada’s boreal forest to be a net source of C for several further) [54] will eventually erase short-lived cooling decades [37]. Although the implications for C dynamics effects. have not been studied extensively [41], insect disturbance was responsible for a greater loss of stored C than was fire Recommendations to manage biodiversity and carbon from Canadian forests in the late 20th century [42], and the retention simultaneously estimated annual C release due to the current mountain Considering that boreal regions are at latitudes where pine beetle outbreak in western Canada is 50% more than climate warming will be globally most profound [4],itis rates attributable to fires during even the most severe fire our opinion that current practices of boreal forest man- years [40]. agement (Box 1) are inadequate to deal with the pace and magnitude of expected changes [55]. The essential role of Role of timber harvest boreal forests in C sequestration itself is strong justifica- The extent of boreal forest harvest varies widely by region tion to create large forest reserves [44]. Such large forest (see above) but, even in areas with high harvest rates, the reserves are possible in Canadian and Russian boreal long-term impact of logging on C stores may be limited if forests, and we argue that these countries in particular there is regeneration to forested stands [43], particularly have a moral and global responsibility to create such because much of the C removed is not released immedi- reserves. However, while old-growth forests are important ately but is stored in various commercial products or ends for C sequestration [44], reserves should be sufficiently up in landfills. Following harvest, there can be an initial large to accommodate a natural disturbance regime which, release of C depending on silvicultural practice, rotation in turn, will maintain a wide range of seral stages to length and harvest-related damage, but sequestration maximize the area available for habitat-specialist species eventually returns to original rates as new stands grow [56]. This can be achieved by incorporating stand structure [44]. Although there is disagreement [e.g. 45], soil C gener- and complexity into reserve-design algorithms [57]. Large ally appears to remain approximately constant [46]. How- reserves in the boreal forest are also needed as ‘living ever, above-ground stores decrease in managed landscapes laboratories’ to understand the effects of natural disturb- for 10 years after harvest, after which the site becomes a ances, and as a ‘natural capital bank’ against the unfore- net C sink as new stands grow [47]. Overall, the effects of seen [58]. To maximize C sequestration, increasing the natural disturbance such as fire on C sequestration are scale of reforestation in heavily disturbed Fennoscandia probably more important than the impact of management and restricting the massive deforestation and fragmenta- interventions such as extending harvest rotation, enhan- tion in Russia arising from timber harvest, mining, hydro- cing regeneration, or increasing stocking densities [37].In power dam construction, and the development of oil and fact, the opening up of forests for access to harvest sites in gas should be a top priority for forest managers [6]. One Russia may have led to a higher frequency of fire [17], possible way to offset the lost economic opportunities from indicating that landscape management can have broader curtailing industrial exploitation is to extend ‘reducing implications for C storage. In other words, the fragmenta- carbon emissions from deforestation and forest degra- tion resulting from the harvest and management of timber dation’ (REDD) credits [59] to these regions.

545 Author's personal copy

Opinion Trends in Ecology and Evolution Vol.24 No.10

Box 1. Principles of current boreal forest management for ill-equipped to adapt to the effects of climate warming biodiversity conservation on natural fire patterns, and a vestigial appreciation of experimental adaptive management will quickly compro- Boreal forest management is driven by local context. Live tree retention is widely practised to achieve various outcomes (re- mise this relatively intact, but latently threatened [61], seeding potential, wildlife use) [62], yet this technique may not be biodiversity haven. Unlike many of the world’s highly adequate to conserve many taxa. Insects, cryptogams and fungi are degraded ecosystems, we have the opportunity to preserve essential for the decomposition of woody debris and nutrient boreal forests and the species they harbour while main- cycling, but saproxylic species richness correlates positively with taining an effective C sink. Ideally, civil society, econom- the amount of dead wood retained [63–65] and species have already disappeared from Fennoscandia due to the paucity of dead wood in ists, social scientists, biologists, policymakers and managed forests [66]. Dead and decaying trees may also be politicians must work more closely to manage the boreal important for maintaining birds through the provision of nesting forest effectively. habitat [67]. Likewise, trees are retained as corridors to facilitate animal movement among fragments, but it is unclear if these are Acknowledgements effective for population persistence at broader scales [68,69]. NSS was supported by a Sarah and Daniel Hardy Visiting Fellowship at In Fennoscandia, small patches containing threatened species are Harvard University while this manuscript was prepared. We thank F. preserved in managed forests [70,71], but it is uncertain if these Achard (Institute for Environmental Sustainability, Joint Research maintain vital ecosystem functions such as seed dispersal [69] Centre of the European Commission) and C. Pollock (IUCN Species critical for forest regeneration [72]. Similar management ap- Programme) for assistance in accessing data. proaches more broadly applied include the retention of riparian buffer strips in logged forests to protect wetlands and biodiversity, but there is a greater need to be able to vary the width of these strips Appendix A. Supplementary data depending on wetland position, watershed connectivity, hydrology Supplementary data associated with this article can be and biodiversity requirements [73,74]. Likewise, because some rural found, in the online version, at doi:10.1016/j.tree.2009. communities in the boreal region rely on non-timber resources such 03.019. as bushmeat, mushrooms, berries and firewood, institutions and logging companies have begun to devise plans to manage forests for diversity beyond the timber products they provide [58]. References The dominance of natural fire in shaping boreal forests [58] has 1 Bradshaw, C.J.A. et al. (2009) Tropical turmoil – a biodiversity tragedy been used to justify the practice of clear-felling [68], and the in progress. Front. Ecol. Evol. 7, 79–87 application of fire itself is used as a management tool [75]. However, 2 Bonan, G.B. (2008) Forests and climate change: forcings, feedbacks, this strategy can be problematic because natural fire regimes can be and the climate benefits of forests. Science 320, 1444–1449 difficult to emulate [69] principally because the high spatio-temporal 3 Warkentin, I.G. and Sodhi, N.S. (2008) Financing tropical forest variability makes single-prescriptions unrealistic [68]. Additionally, preservation. Science 320, 874 the uncommon practice of repeated burning might be needed for the 4 Ruckstuhl, K.E. et al. (2008) Introduction. The boreal forest and global germination of seed banks [67,68]. Other types of disturbances such change. Proc. R. Soc. Lond. B. Biol. Sci 363, 2245–2249 as insect outbreaks, pathogens, windstorms, droughts and floods [68] 5 Soja, A.J. et al. (2007) Climate-induced boreal forest change: and their complex synergies with fire are generally intractable to predictions versus current observations. Global Planet. Change 56, emulate as management tools. The mushrooming certification 274–296 schemes available for sustainable boreal forestry [58,76] driven by 6 Achard, F. et al. (2006) Areas of rapid forest-cover change in boreal demands from enlightened consumers and environmental activists Eurasia. For. Ecol. Manage. 237, 322–334 might only be partially effective [77] to maintain biodiversity values, 7 Smith, W. and Lee, P., eds (2000) Canada’s Forests at a Crossroads: An but more research is needed to test their efficacy. Assessment in the Year 2000, World Resources Institute 8 Mollicone, D. et al. (2006) Human role in Russian wild fires. Nature 440, 436–437 Given the dominance of natural disturbances in driving 9 Zhuang, Q. et al. (2006) CO2 and CH4 exchanges between land boreal ecosystem dynamics [17], greater emphasis on ecosystems and the atmosphere in northern high latitudes over the 21st century. Geophys. Res. Lett. 33, L17403 managing expected modification of these disturbances 10 Potapov, P. et al. (2008) Combining MODIS and Landsat imagery to resulting from climate change can be more effective than estimate and map boreal forest cover loss. Remote Sens. Environ. 112, concentrating exclusively on managing harvest regimes, 3708–3719 from both a C sequestration and biodiversity conservation 11 Olson, D.M. et al. (2001) Terrestrial ecoregions of the world: a new map perspective. The sheer extent of the boreal forests within of life on Earth. Bioscience 51, 933–938 12 Achard, F., et al., (2005) Identification of Hot Spot Areas of Forest Russia and Canada, combined with the large ranges of Cover Changes in Boreal Eurasia (EUR 21684 EN), European many shared taxa [4], are most likely responsible for the Commission relative low frequency of endangerment observed (Online 13 FAO (2006) Global Forest Assessment 2005. FAO Forestry Paper 147, supplementary material Tables S3–S5). Thus, continued 1-320 14 Greenpeace (2006) Roadmap to Recovery. The World’s Last Intact fragmentation from natural and human-driven processes Forest Landscapes, Greenpeace International is perhaps the greatest future concern for species conser- 15 Brooks, T.M. et al. (2002) Habitat loss and extinction in the hotspots of vation there. While fragmentation remains a clear threat biodiversity. Conserv. Biol. 16, 909–923 [63], forest management must not only consider fragmen- 16 United Nations (2000) Forest Resources of Europe, CIS, North tation, it must also attempt to avoid creating large stands America, Australia, Japan and New Zealand (industrialised temperate boreal countries). UN-ECE/FAO Contribution to the of even-aged trees [60], maximize connectivity of existing Global Forest Resources Assessment 2000. (ECE/TIM/SP/17), United fragments, and consider the implications of the storage and Nations release of C at regional and continental scales. Clearly, 17 Achard, F. et al. (2008) The effect of climate anomalies and human there must also be better management in Russia to reduce ignition factor on wildfires in Russian boreal forests. Philos. Trans. R. the frequency of human-caused fires. Soc. Lond., B 363, 2331–2339 18 Sukhinin, A.I. et al. (2004) AVHRR-based mapping of fires in Russia: It is our opinion that the status quo of current rates new products for fire management and carbon cycle studies. Remote of forest fragmentation, stale management practices Sens. Environ. 93, 546–564

546 Author's personal copy

Opinion Trends in Ecology and Evolution Vol.24 No.10

19 Mouillot, F. and Field, C.B. (2005) Fire history and the global carbon 45 Ter-Mikaelian, M.T. et al. (2008) Fact and fantasy about forest carbon. budget: a 18 Â 18 ﬁre history reconstruction for the 20th century. Glob. For. Chron. 84, 166–171 Change Biol. 11, 398–420 46 Fredeen, A. et al. (2005) Comparison of coniferous forest carbon stocks 20 Mo¨nkko¨nen, M. and Viro, P. (1997) Taxonomic diversity of the between old-growth and young second-growth forests on two soil types terrestrial bird and mammal fauna in temperate and boreal biomes in central British Columbia, Canada. Can. J. For. Res. 35, 1411– of the Northern Hemisphere. J. Biogeogr. 24, 603–612 1421 21 IPCC (2007) Climate Change 2007: Synthesis Report. Contribution of 47 Fredeen, A. et al. (2007) When do replanted sub-boreal clearcuts

Working Groups I, II and III to the Fourth Assessment Report of the become net sinks for CO2? For. Ecol. Manage. 239, 210–216 Intergovernmental Panel on Climate Change 104, Intergovernmental 48 Wirth, C. (2005) Fire regime and tree diversity in boreal forests: Panel on Climate Change implications for the carbon cycle. In Forest Diversity and Function 22 Stephens, B.B. et al. (2007) Weak northern and strong tropical land Temperate and Boreal Systems Ecological Studies (Scherer-Lorenzen,

carbon uptake from vertical profiles of atmospheric CO2. Science 316, M. et al., eds), pp. 309–346, Springer 1732–1735 49 Flannigan, M. et al. (2008) Impacts of climate change on fire activity 23 Intergovernmental Panel on Climate Change (2001) Climate Change and fire management in the circumboreal forest. Glob. Clim. Change 2001: The Scientific Basis, Cambridge University Press 14, 1–12 24 Kelloma¨ki, S. et al. (2008) Sensitivity of managed boreal forests in 50 Betts, R.A. (2000) Offset of the potential carbon sink from boreal Finland to climate change, with implications for adaptive forestation by decreases in surface albedo. Nature 408, 187–190 management.. Philos. Trans. R. Soc. Lond., B 363, 2341–2351 51 Bala, G. et al. (2007) Combined climate and carbon-cycle effects of 25 Chen, J.M. et al. (2006) Boreal ecosystems sequester more carbon in large-scale deforestation. Proc. Natl. Acad. Sci. U. S. A. 104, 6550– warmer years. Geophys. Res. Lett. 33, L10803 6555

26 Euskirchen, E.S. et al. (2006) Importance of recent shifts in soil thermal 52 Brovkin, V. et al. (2005) Role of land cover changes for atmospheric CO2 dynamics on growing season length, productivity, and carbon increase and climate change during the last 150 years. Glob. Change sequestration in terrestrial high-latitude ecosystems. Glob. Change Biol. 10, 1253–1266 Biol. 12, 731–750 53 Snyder, P. et al. (2004) Evaluating the inﬂuence of different vegetation

27 Hyvo¨nen, R. et al. (2007) The likely impact of elevated [CO2], nitrogen biomes on the global climate. Clim. Dyn. 23, 279–302 deposition, increased temperature and management on carbon 54 Serreze, M.C. and Francis, J.A. (2006) The Arctic on the fast track of sequestration in temperate and boreal forest ecosystems: a change. Weather 61, 65–69 literature review. New Phytol. 173, 463–480 55 Ogden, A.E. and Innes, J. (2007) Incorporating climate change 28 Canadell, J.G. et al. (2007) Contributions to accelerating atmospheric adaptation considerations into forest management planning in the

CO2 growth from economic activity, carbon intensity, and efficiency of boreal forest. Int. For. Rev. 9, 713–733 natural sinks. Proc. Natl. Acad. Sci. U. S. A. 104, 18866–18870 56 Berg, A. et al. (1994) Threatened plant, animal and fungus species in 29 Houghton, R.A. et al. (2007) Mapping Russian forest biomass with data Swedish forests: distribution and habitat associations. Conserv. Biol. 8, from satellites and forest inventories. Environ. Res. Lett. 2, 045032 718–731 30 Goodale, C.L. et al. (2002) Forest carbon sinks in the northern 57 Williams, J.C. et al. (2004) Using mathematical optimization models to hemisphere. Ecol. Appl. 12, 891–899 design nature reserves. Front. Ecol. Evol. 2, 98–105 31 Fang, J. et al. (2006) Overestimated biomass carbon pools of the 58 Burton, P.J. et al. (2006) Sustainable management of Canada’s boreal northern mid- and high latitude forests. Clim. Change 74, 355–368 forests: progress and prospects. Ecoscience 13, 234–248 32 Bond-Lamberty, B. et al. (2007) Fire as the dominant driver of central 59 Miles, L. and Kapos, V. (2008) Reducing greenhouse gas emissions from Canadian boreal forest carbon balance. Nature 450, 89–92 deforestation and forest degradation: global land-use implications. 33 Crevoisier, C. et al. (2007) Drivers of fire in the boreal forests: data Science 320, 1454–1455 constrained design of a prognostic model of burned area for use in 60 Bergeron, Y. et al. (2001) Natural fire frequency for the eastern dynamic global vegetation models. Geophys. Res. Lett. 112, D24112 Canadian boreal forest: consequences for sustainable forestry. Can. 34 Flannigan, M.D. et al. (2005) Future area burned in Canada. Clim. J. For. Res. 31, 384–391 Change 72, 1–16 61 Cardillo, M. et al. (2006) Latent extinction risk and the future 35 Kang, S. et al. (2006) Simulating effects of fire disturbance and climate battlegrounds of mammal conservation. Proc. Natl. Acad. Sci. change on boreal forest productivity and evapotranspiration. Sci. Total U. S. A. 103, 4157–4161 Environ. 362, 85–102 62 Franklin, J.F. et al. (1997) Alternative silviculture approaches to timber 36 Balshi, M.S. et al. (2007) The role of historical fire disturbance in the harvesting: variable retention systems. In Creating Forestry for the 21st carbon dynamics of the pan-boreal region: a process-based analysis. J. Century (Kohn, K.A. and Franklin, J.F., eds), pp. 111–139, Island Press Geophys. Res. 112, G02029 63 Martikainen, P. et al. (2000) Species richness of Coleoptera in mature 37 Kurz, W.A. et al. (2008) Risk of natural disturbances makes future and old-growth boreal forests in southern Finland. Biol. Conserv. 94, contribution of Canada’s forests to the global carbon cycle highly 199–209 uncertain. Proc. Natl. Acad. Sci. U. S. A. 105, 1551–1555 64 Kourki, J. et al. (2001) Forest fragmentation in Fennoscandia: linking 38 Candau, J-N. and Fleming, R.A. (2005) Landscape-scale spatial habitat requirements of wood-associated threatened species to distribution of spruce budworm defoliation in relation to bioclimatic landscape and habitat changes. Scand. J. For. Res. 3, 27–37 conditions. Can. J. For. Res. 35, 2218–2232 65 Ehnstro¨m, B. (2001) Leaving dead wood for insects in boreal forests - 39 Seidl, R. et al. (2008) Impact of bark beetle (Ips typographus L.) suggestions for the future. Scand. J. For. Res. 3, 91–99 disturbance on timber production and carbon sequestration in 66 Sittonen, J. and Martikainen, P. (1994) Occurrence of rare and different management strategies under climate change. For. Ecol. threatened insects on decaying Populus tremula: a comparison Manage 256, 209–220 between Finnish and Russian Karelia. Scand. J. For. Res. 9, 185–191 40 Kurz, W.A. et al. (2008) Mountain pine beetle and forest carbon 67 Spence, J. (2001) The new boreal forestry: adjusting timber feedback to climate change. Nature 452, 987–990 management to accommodate biodiversity. Trends Ecol. Evol. 16, 41 Volney, W.J.A. and Fleming, R.A. (2000) Climate change and impacts 591–593 of boreal forest insects. Agric. Ecosyst. Environ. 82, 283–294 68 Simberloff, D. (2001) Management of boreal forest biodiversity – a view 42 Kurz, W.A. and Apps, M.J. (1999) A 70-year retrospective analysis of from the outside. Scand. J. For. Res. 16 (Suppl. 3), 105–118 carbon fluxes in the Canadian forest sector. Ecol. Appl. 9, 526–547 69 Niemela¨,J.et al. (2001) Concluding remarks – finding ways to 43 Prentice, I.C. et al. (2001) The carbon cycle and atmospheric carbon integrate timber production and biodiversity in Fennoscandian dioxide. In Climate Change 2001: The Scientific Basis. Contribution of forestry. Scand. J. For. Res. 16 (Suppl. 3), 119–123 Working Group I to the Third Assessment Report of the 70 Hansson, L. (2001) Key habitats in Swedish managed forests. Scand. J. Intergovernmental Panel on Climate Change (Houghton, J.T. et al., For. Res. 3, 52–61 eds), pp. 183–237, Cambridge University Press 71 Raivio, S. et al. (2001) Science and management of boreal forest 44 Luyssaert, S. et al. (2008) Old-growth forests as global carbon sinks. diversity – a forest industries’ view. Scand. J. For. Res. 16 (Suppl. 1), Nature 455, 213–215 100–105

547 Author's personal copy

Opinion Trends in Ecology and Evolution Vol.24 No.10

72 Ordonez, J.L. and Retana, J. (2004) Early reduction of post-fire Sustainable Management of the Boreal Forest (Burton, P.J. et al., recruitment of Pinus nigra by post-dispersal seed predation in eds), pp. 857–892, NRC Research Press different time-since-fire habitats. Ecography 27, 449–458 77 Sverdrup-Thygeson, A. et al. (2008) A comparison of biodiversity values 73 Macdonald, S.E. et al. (2006) Is forest close to lakes ecologically unique? in boreal forest regeneration areas before and after forest certification. Analysis of vegetation, small mammals, amphibians, and songbirds. Scand. J. For. Res. 23, 236–243 For. Ecol. Manage. 223, 1–17 78 Bartalev, S.A. et al. (2003) A new SPOT4-VEGETATION derived 74 Devito, K.J. et al. (2000) Landscape controls on phosphorous loading in land cover map of Northern Eurasia. Int. J. Remote Sens. 24, 1977– boreal lakes: implications for the potential impacts on forest 1982 harvesting. Can. J. Fish. Aquat. Sci. 57, 1977–1984 79 Latifovic, R. et al. (2004) Land cover mapping of North and Central 75 Hagner, S. (1999) Forest Management in Temperate and Boreal Forests: America-Global Land Cover 2000. Remote Sens. Environ. 89, 116– Current Practices and the Scope for Implementing Sustainable Forest 127 Management, United Nations 80 Hansen, M.C. et al. (2003) Global percent tree cover at a spatial 76 Duinker, P.N. and Trevisan, L.M. (2003) Adaptive management: resolution of 500 meters: first results of the MODIS Vegetation progress and prospects for Canadian forests. In Towards Continuous Fields algorithm. Earth Interact. 7, 1–15

548